2.1 Why is Equity Relevant?

2.1.1 It may be thought that equity is not a relevant issue where shareholders and policyholders share profits in a fixed ratio: the 90:10 division leads to incentives to increase profits so that both parties gain. With-profits policyholders can certainly benefit from management operating the fund in this way. They also benefit from guarantees that apply; but this can mean that losses are not shared in the same way as profits, and can lead to potential conflicts between shareholders’ and policyholders’ interests.

2.1.2 It may be argued that questions of equity do not arise if a policy is sold on a basis that is set out, and is then operated accordingly. However, with-profits life contracts are not well-defined, not well-understood by policyholders (FSA, 2002a) and leave significant discretion to management. Policyholders are therefore in a weak position (Sandler, Reference Sandler2002) and there is concern that discretion may be exercised in a way that disadvantages policyholders (FSA, 2003). Although insurers have, since 2004, issued Principles and Practices of Financial Management (PPFM) documents, which help define how insurers operate their business, considerable management discretion still remains. There is, therefore, potential for inequity to arise.

2.1.3 The FSA recognises this, and the outcome of its with-profits review was changes to its rules, including new rules on governance arrangements and the conduct of business. These were additional to the FSA's general principles, which included the requirements to treat customers fairly and to manage conflicts of interest fairly, both being important issues for proprietary with-profits funds.

2.2 The Meaning of Fairness

2.2.1 The FSA (2002b) said that fairness is a powerful concept about which people have deeply intuitive but varied notions, with natural justice, ethics and morality all potentially playing a part. Honesty, openness, transparency and good faith may be indicators that a firm is acting fairly.

2.2.2 The FSA's “Treating Customers Fairly” (TCF) initiative led the regulator to express the desired outcomes of TCF, which included, inter alia, customers receiving the product performance they have been led to believe from firms they deal with (FSA, 2006:5). An actuarial working party (Thompson et al., Reference Thompson, Bright, Cummings and Shah2002) felt that fairness included “honouring representations, assurances and guarantees which create legitimate expectations”. The author suggests these perspectives are consistent with a legal approach: firms should fulfil their obligations under contracts, and doing otherwise is unfair. The issue is to establish what are the obligations under imprecisely defined with-profits contracts. This approach does not rule out other approaches to fairness but the author believes it is helpful in deciding on the fairness of decisions that with-profits insurers take.

2.3 A Legal Approach: Implied Terms

2.3.1 The question is approached by considering how courts decide on parties’ rights and obligations under contracts, and apply this to with-profits contracts. Courts have, in a number of cases, intervened in disagreements about the bonus rates declared on with-profits policies (O'Brien, Reference O'Brien2004). The courts’ perspective is to interpret contracts and, in some circumstances, to imply terms into contracts. Relevant for this paper are ‘implied terms in fact’ and ‘implied terms in law’.

2.3.2 Terms that are implied in fact are those that rely on the facts of the case in question: see Appendix B. While it may be said that courts imply terms that they believe reflect the intention of the parties, Rakoff (Reference Rakoff1995:192) argues that this is not always the case: “intention is often the fiction”.

2.3.2 Terms that are implied in law are those that are implied into all contracts of a particular type because of the nature of the contract, e.g. master/servant, employer/employee: see Appendix B. The courts imply a term as a result of the particular relationship between the parties, being a default rule which will apply unless specifically excluded. Terms implied in law do not depend on the parties’ intentions; and are predicated on broader considerations of public policy (Phang, Reference Phang1993).

2.4 Implied Terms: The Case of With-Profits Contracts

2.4.1 If a court were considering a typical with-profits contract, would it be appropriate to decide on terms implied by law? There are two specific tests (see Appendix B). The first is that the contract must be of a defined type. With-profits contracts are defined in the rules of the FSA. The contracts vary in type (covering endowment assurances, whole life policies, pensions and so on) and they vary in the detail of their terms and conditions, including participation rights; in particular, demutualised insurers may operate in accordance with some specific conditions. However, given that terms in law can be applied to employment contracts, which is a wide category containing a significant variety, the author's view is that the ‘defined type’ test is met for with-profits contracts.

2.4.2 Second is the test referred to as ‘necessity’. This does not mean that the contract cannot be operated without the implied term; courts have a broad view of what necessity means. In particular, the criterion of necessity is broader for terms implied by law than for terms implied in fact (Phang, Reference Phang1998): if courts feel that issues of social policy, fairness and justice make it necessary to have terms implied in law, they will decide accordingly.

2.4.3 Koffman & Macdonald (Reference Koffman and Macdonald2007) say that an imbalance in the relative bargaining power of the parties may be grounds for implication of terms in law; this is a way of addressing issues of social policy and fairness. There is a similarity with the way in which unfair contract terms legislation renders contract terms inapplicable if there is an imbalance in parties’ rights and obligations and which act to customers’ detriment. In other words, terms implied by law may be similar to negating terms that would be judged unfair.

2.4.4 With-profits contracts are characterised by an imbalance in the relative bargaining power of the parties; the contract is complex, and the insurer as the decision-making body has discretion in operating the business. With-profits contracts raise issues of reasonableness, fairness and the balancing of competing issues. The author therefore suggests that with-profits contracts are of a nature that means implication of terms by law is feasible.

2.4.5 The author would not accept arguments that implication of terms is unnecessary because competition effectively requires insurers to act fairly; the following points take into account Paragon Finance v Nash (see appendix B):

• – competition in the retail financial services market does not work well (FSA, 2011a); in particular, the with-profits market does not work satisfactorily, with consumers in a weak position to understand the complexities of with-profits (Sandler, Reference Sandler2002); and the FSA's (2011b) analysis drew attention to problems in the operation of the with-profits market which can result in unfair outcomes for with-profits policyholders;

• – there is a declining number of insurers offering with-profits business and the market is becoming increasingly concentrated, with many funds now closed to new business (O'Brien, Reference O'Brien2009a);

• – policyholders incur costs in moving to an alternative provider, such as surrender penalties, the need to pay for acquisition expenses included in the new provider's pricing, and non-monetary costs; and

• – policyholders may be unable to find a new provider if the insured's health or other circumstances have changed or may be disadvantaged if there have been changes in the market, such as the products that are available and how they are priced.

2.4.6 Further, the author would not accept arguments that implication of terms is unnecessary because of existing regulation; the following points again take into account Paragon Finance v Nash:

• – The existence of regulation is evidence that firms may, if unconstrained, act in a way that is regarded as undesirable, but regulation itself cannot guarantee that firms’ actions are desirable;

• – regulation may be affected by political considerations and the influence of stakeholders (Meier, Reference Meier1991); and

• – Equitable Life v Hyman illustrated that consideration by the courts of implied terms may lead to a different outcome from that resulting from regulation.

2.4.7 A separate argument is that implication of terms in law can lead to uncertainty and unpredictability; and that the courts are effectively determining policy, a role better fitted to the legislature. Phang (Reference Phang1990) says the arguments are finely balanced. However, implication of terms in law has become an accepted form of court practice.

2.4.8 In the case of life insurance, it could be argued that as there is a regulator, it should determine policy rather than the courts imply terms in law. The regulator uses consultation and cost-benefit analysis for policy changes, which might means the outcomes are more robust; and courts may lack expertise. On the other hand, the FSA has, in the past, made concessions to the industry (see sections 5.4 and 5.5), which raises questions about its ability to balance competing policy considerations.

2.4.9 The distinction between implied terms in fact and in law is important. Thompson et al. (Reference Thompson and Ayliffe2009) suggested five principles applicable to with-profits funds, which they indicated could have legal effect as implied terms. This drew a response from Aviva UK Life (2009a). Aviva's legal advisers’ view was that if terms are to be implied, it must be strictly necessary to do this in order to properly construe the contract, a standard they regarded as not being met. They went on to say that if it is not strictly necessary to imply terms, the court will look to the presumed intention of the parties. They quote Lord Wilberforce in Liverpool City Council v Irwin: the courts must be “spelling out what both parties know and would, if asked, unhesitatingly agree to be part of the bargain”. They considered that both parties’ (and particularly not the insurer's) intentions could not be presumed to encompass the five principles (which are not examined in this paper).

2.4.10 However, the test of necessity is interpreted by the courts in a broad way. The issue is how to properly construe a contract where the insurer has greater knowledge and discretion and has issued a contract with standardised terms. The case of Equitable Life v Hyman illustrates the point that although it was feasible to operate the contract on some basis, implication of terms was felt to be necessary to achieve an outcome that reflected the contract. Further, where terms are implied in law, the criterion of necessity takes into account broader policy considerations (Phang, Reference Phang1998).

2.4.11 The presumed intentions of the parties are also irrelevant when the courts imply terms in law. In Liverpool City Council v Irwin, Lord Wilberforce indicated that there were a number of reasons for implying terms: Aviva UK Life (2009a) refers to one (above) but omits the different justification used in that case, which was one of implication in law (see Appendix B).

2.4.12 This paper therefore regards it as appropriate to consider how some specific issues in with-profits business would be viewed in the light of how the courts might imply terms in law. However, a court may use implied terms in fact to derive the same outcome, given that courts do not always require both parties to agree on the implication.

2.4.13 The courts would not imply terms that are inconsistent with express terms of a contract (assuming such express terms are themselves not unfair: see section 2.5).

2.5 Unfair Terms

2.5.1 The potential for contracts to be unfair has been recognised in legal literature and by the Unfair Terms in Consumer Contracts Regulations 1999 (UTCCR), to which life insurers are subject: see Appendix C. Regulation 5(1) contains the following test of fairness:

2.5.2 What are sometimes called core terms, i.e. those which describe the main subject matter of the contract or concern the adequacy of the price, are exempted from the Regulations, provided they are expressed in plain and intelligible language.

2.5.3 There are therefore four requirements for a term to be unfair. First is that the term is not individually negotiated, which will generally be satisfied for with-profits contracts.

2.5.4 Second is that there is a significant imbalance in the parties’ rights and obligations. Kennedy (Reference Kennedy1994) indicates that charges of imbalance are inevitable in insurance contracts. For with-profits business the following points apply (this section draws on FSA, 2003):

• – The firm has decision-making powers, while consumers have little, if any, influence;

• – The firm has discretion in how it manages the business, including the payouts under policies;

• – The firm has rights that have a long-term perspective, with shareholder value encompassing cashflows over many years into the future, whereas the customer's perspective may not stretch beyond the duration of his or her policy; and

• – A dissatisfied customer will incur costs and inconvenience in moving to another insurer.

2.5.5 In exercising rights, consumers are in a weak position:

• – The firm has more information about the contract than the consumer;

• – The contracts are complex, and the consumer may find it difficult to understand them;

• – The consumer may not have the time, expertise and financial resources to challenge the firm regarding the contract, whereas the firm can more easily challenge the consumer; and

• – The degree of competition is less after the policy is in force, and the insurer may be closed to new business.

2.5.6 In particular, firms make decisions about the business in the areas discussed in section 5, illustrating the imbalance in the parties’ rights.

2.5.7 The third requirement is that the imbalance is “contrary to the requirements of good faith”. Appendix C reviews this point and concludes that if the other conditions for unfairness are present, this automatically means there is unfairness.

2.5.8 The fourth requirement is that there is consumer detriment, which is considered later in specific situations.

2.5.9 More recent rules are the Consumer Protection from Unfair Trading Regulations 2008 (CPUTR), derived from the EU Unfair Commercial Practices Directive 2005 (see Cartwright, Reference Cartwright2010). The FSA has taken account of these regulations in drawing up its rules.

3.1 Fairness in With-Profits Contracts

3.1.1 This paper adopts the position that fairness requires insurers to deliver the obligations in their contracts, but not unfair obligations. In other words, an insurer acting fairly will:

• – comply with its obligations under the contract;

• – comply with any terms that would be implied by the courts;

• – not seek to impose any terms that would be unfair under the Unfair Terms in Consumer Contracts Regulations 1999; and

• – not operate any commercial practices that are unfair in accordance with the Consumer Protection from Unfair Trading Regulations 2008.

3.1.2 As CPUTR is largely similar to existing regulations affecting insurance, this paper concentrates on the previous points.

3.1.3 The suggestion above has similarities to Tuley's (Reference Tuley2009) suggestion that fairness involves a firm not performing a contract in a way which undermines an important part of the contractual bargain between the parties, with constraints on the exercise of discretion. It is appreciated that others may set further requirements for fairness.

3.1.4 Implying terms takes into account the commercial setting of the agreement. For with-profits business the setting is:

• – the surplus is calculated as the excess of assets over liabilities of the with-profits fund as a whole: this has been so from the beginnings of with-profits business;

• – there are inter-generational transfers of surplus: not all surplus is distributed immediately, but some is typically held back, again this has been customary from when with-profits business began. Firms may later distribute profits that arose from a previous generation of policyholders but which were held back for some reason;

• – not less than 90% of the distributed surplus is allocated to policyholders. This implies that shareholders will not receive a distribution of profits unless policyholders receive a distribution (of nine times the amount); and this 90:10 formula applies not only to the distribution of profits representing profits within asset shares but also to profits arising from inter-generational transfers.

3.1.5 The author refers to these as the three basic with-profits principles. Some contracts may have express terms to the contrary but these are not typical.

3.1.6 In support of the first principle, FSA stated, with-profits policyholders “have an interest in the whole with-profits fund and in every part of it, which derives from the fact that the with-profits fund is a single, undivided fund, from which any particular assets could be used to meet the fund's contractual obligations in respect of a with-profits policy written into that fund” (FSA, 2009:9, and confirmed in FSA, 2011b:9).

3.1.7 As regards the third principle, the 90:10 formula has been criticised as arbitrary (Sandler, Reference Sandler2002) and as giving excessive returns to shareholders (Wallace, Reference Wallace1973; Redington, Reference Redington1981). Nevertheless, this paper takes it as given: policyholders accept the 90% share of profits that the firm allocates to them; in essence, the 10% of surplus forgone by policyholders is part of the price of the product.

3.1.8 There are a number of practices and methods that insurers use to run the business. These include methods of determining and managing asset shares and, separately, the inherited estate (the excess of realistic assets over realistic liabilities). However, such practices and methods are subsidiary to the three basic with-profits principles, with which they are expected to be consistent, and the basic principles do not refer specifically to them.

3.2 Procedural Fairness

3.2.1There is an argument that with-profits contracts involve procedural unfairness and are therefore unfair. This argument is considered in Appendix C, which suggests it may not be valid.

3.3 Fairness and Discretion

3.3.1 A point concerning discretion can be suggested at the outset. This is that it is implied that an insurer, in exercising discretion, should not act dishonestly, for an improper purpose, capriciously or arbitrarily, or would not exercise its discretion in a way that no reasonable insurer, acting reasonably, would do. This is consistent with the judgement in Paragon Finance v Nash, and the author would suggest it would be an implied term in law if a court determined what were implied terms in law (Appendix A lists the implied terms derived in this paper).

4.1 Introduction

4.1.1 The FSA has been active recently in a number of areas involving policyholder/shareholder issues, notably when deciding that mis-selling compensation cannot be charged to asset shares or the inherited estate of a proprietary with-profits insurer (FSA, 2009). Section 4 refers to other issues referred to in the literature discussed in the literature but where, in the author's view, there are no significant issues now requiring resolution. Sections 5 and 6 consider issues where differences of view remain.

4.2 With-Profits Business with a Lower than 10% Allocation to Shareholders

4.2.1 Some insurers have developed products where they specified that the proportion of distributed profits payable to shareholders is less than 10%. This may reflect competitive pressures. However, some of those firms still calculated the overall transfer to shareholders as 1/9^th of the value of bonuses. If a policy is written on the basis that 5% (say) of surplus is distributable to shareholders, this means that the remaining 5% is paid to the shareholders from the inherited estate.

4.2.2 This, however, contravenes regulatory requirements. KPMG (2009) indicated, “It is our understanding that FSA no longer permits business to be written so that the estate can be eroded in such a manner, and therefore such business should no longer be being written” (p. 10). This reflects the FSA's (2009) view that with-profits policyholders have a contingent interest in the entire value of a with-profits fund, which includes any inherited estate (contingent on the distributions that are made). It is therefore to customers’ detriment, and hence unfair, if firms adversely affected that contingent interest, as was happening.

4.2.3 The inherited estate is, of course, surplus. To distribute part of it to shareholders alone is inconsistent with the basic with-profits principles, which require at least 90% of any distribution (from the inherited estate or other surplus) to be to policyholders. The practice was therefore, in the author's view, inconsistent with the implied terms of with-profits contracts.

4.3 Market Value Reductions

4.3.1 In 2007, the FSA was concerned that with-profits policyholders as a class should not be disadvantaged compared with shareholders as a result of the operation of market value reductions (MVRs) on unitised with-profits policies (FSA, 2007a).

4.3.2 If a firm declared reversionary bonuses, and then used an MVR, the effect of which was to claw back some of those bonuses, shareholders would gain unduly if they received 10% of the surplus giving rise to the reversionary bonus but if the whole of the MVR was attributed to policyholders. In other words, policyholders could receive 90% of the upside but bear 100% of the downside, an issue raised by O'Brien (Reference O'Brien2003).

4.3.3 The FSA changed its rules in 2007 to prevent this practice, the intention being to ensure with-profits firms “do not indulge in practices that could lead to a declared distribution ratio drifting in favour of shareholders and so against the interests of policyholders at the point of an annual distribution and over time” (FSA, 2007a). A slight modification of the rule was proposed by the FSA (2011b).

4.3.4 This section concludes by drawing attention to the guidance in the FSA's COBS 20.2.1, which asks firms to “give careful consideration to any aspect of its operating practice that has a bearing on the interests of its with-profits policyholders to ensure that it does not lead to an undisclosed, or unfair, benefit to shareholders.” The FSA (2011b) proposed that a slightly modified version of this be changed from guidance to a rule.

4.4. Staff Pension Schemes

4.4.1 Which? (2008) complained that the FSA's rules permitted proprietary firms to use the inherited estate to pay off a deficit in a staff pension scheme. However, the profits of a life insurer are after expenses, which include staff pension scheme costs as applicable to the insurer. It is therefore acceptable (in principle, at least) to charge such expenses to asset shares. If the expenses are charged to the inherited estate then the cost is borne by those who benefit from the inherited estate, i.e. both policyholders and shareholders. There does not therefore appear to be a specific advantage to shareholders of charging staff pension scheme costs to with-profits funds.

4.4.2 There are some issues concerning calculating staff pension scheme costs (where accounting practice has changed), and the allocation of scheme costs between companies in a group. However, FSA rules should help ensure that appropriate costs are charged to with-profits funds, and the FSA's (2010) review did not find examples of disproportionate amounts of firm's staff pension scheme deficits being charged to a with-profits fund. The author's view is that this is not therefore a current issue requiring resolution and the FSA (2011b) confirmed that it does not plan any changes in this area.

4.5 Issues Concerning With-Profits Funds in Groups

4.5.1 Proprietary with-profits life insurers are typically in a group that also includes one or more life insurers where all profits are allocated to shareholders. This has led to the following situations having been identified:

• – a with-profits fund could write non-profit business but new non-profit business is subsequently written in another company where profits are allocated to shareholders, leading to a loss of goodwill in the with-profits fund (Brindley et al., Reference Brindley, Dumbreck, Thomson and Thompson1990);

• – reinsurance between companies with differing shareholder participation in profits may require attention to avoid unfairness (Brindley et al., Reference Brindley, Cannon, Dumbreck, Thomson and Thompson1993); and

• – cost allocations for projects could be made in ways which are favourable to shareholders (Brindley et al., Reference Brindley, Nowell and Thomson1998; Sandler, Reference Sandler2002).

4.5.2 However, the papers referred to above did not suggest that unfairness had arisen in practice. It could be expected that directors of companies exercising their responsibilities would ensure that the potential conflicts were addressed satisfactorily and the FSA (2010) review did not refer to these issues as a concern. The author's view is that this is not therefore a current issue requiring resolution (however, section 5.3 refers to issues concerning management service companies).

5.1 Management Actions

5.1.1 Life insurers have discretion to change the management of the business, by actions such as changing the investment strategy of the fund, bonus rates, the degree of smoothing and surrender values.

5.1.2 With-profits insurers’ solvency has been depleted in recent years as a result of poor investment returns and other reasons, and one outcome has been management actions such as reducing the degree of smoothing and reducing the equity backing ratio of the fund. Tuley (Reference Tuley2009) suggested these may be inconsistent with TCF. There have been complaints that some insurers are substituting an inferior product for that originally purchased (Treasury Committee, 2004), with policyholders potentially losing some of the benefit of the equity premium. On the other hand, a reduced equity backing ratio lessens the uncertainty about the payout, which some policyholders will regard as beneficial. As regards reducing the degree of smoothing of payouts, Harley & Davies (Reference Harley and Davies2001:41) commented, “If faced with insolvency [companies] would have the option of changing the [smoothing] rules and making more abrupt changes; indeed it would be very odd if companies clung to rules that threatened them with imminent ruin”.

5.1.3 Management actions exemplify the discretion that firms have. The way it is to be exercised is not transparent; policyholders cannot be sure what actions will be taken in what circumstances. There is an imbalance between the rights of firms and customers.

5.1.4 However, is there customer detriment, without which no unfairness arises? One piece of evidence is from the judgment in Paragon Finance v Nash. It was not an implied term that Paragon Finance was unable to take into account its own financial position in deciding what interest rate to charge on a variable interest loan. It had been adversely affected by an increase in the cost of its funding, reflecting the poor default rates it was experiencing, and it was able to take into account such a commercial consideration to increase the interest rates it charged to its customers. It is reasonable to infer that an insurer could take into account factors such as its reducing solvency in deciding to exercise management actions.

5.1.5 A firm may exercise discretion to increase its charges: the FSA's (2005) view was that legitimate increases in the cost of providing a financial service may be a valid reason for this, provided the increases were proportionate. The FSA also commented, concerning protection life insurance contracts with the facility for the insurer to review premiums, that it would not be valid if the insurer increased profitability margins beyond those assumed at the outset (para 4.19).

5.1.6 This is consistent with the view that consumers believe that if firms take actions to increase their profits compared with some reference point, that is unfair (Kahneman et al., Reference Kahneman, Knetsch and Thaler1986).

5.1.7 Say a firm changes its smoothing practice, for which there may be sound financial reasons. There will be an impact on both policyholders and shareholders. One of the effects may be a lower likelihood of the shareholders having to support the with-profits fund financially, and the outcome may be that shareholders are expected to gain at the expense of policyholders. If so, the firm has exercised its discretion in a way that has led to customer detriment, which, in conjunction with the other criteria for unfair terms in section 2.5 is unfair. This might be offset if the shareholders’ share of surplus is reduced to below 10%. To avoid unfairness, the author suggests there is an implied term that with-profits firms will not use their discretion to take management actions that result in an increase in shareholder value at the expense of ‘policyholder value’. The author leaves the question of how shareholder and policyholder value are measured, monitored and regulated to a separate discussion.

5.1.8 The position is complicated as a firm may take management actions to meet the interests of non-profit policyholders, or a group of with-profits policyholders with interests that differ from those of others. Nevertheless, the author would still expect a firm to consider if its actions were leading to increased value for shareholders at the expense of with-profits policyholders.

5.2 Choice of Investments

5.2.1 One approach to the investments of a with-profits fund is to say that they are chosen in the interests of policyholders. However, considering only their interests may lead to risky investments that increase the potential commitment of shareholders. Hence it would be reasonable to say that both shareholders’ and policyholders’ interests in the fund are relevant.

5.2.2 Shareholders may, however, have additional interests in some investments. Sandler (Reference Sandler2002) was concerned that money could be invested in a venture from which the extraction of profits favours shareholders and recommended that with-profits funds should not be used to finance other parts of the provider's business. If those restrictions were inconsistent with the requirements of the First Life Directive, he would recommend disclosure of such investments.

5.2.3 The three basic with-profits principles do not prevent an investment with a company with which the insurer's parent has a strategic connection or interest. However, permitting an investment that was to policyholders’ detriment would be an unfair term. From this, the author deduces that there is an implied term, namely that any such investment should be made only if it is expected to be at least as beneficial to with-profits policyholders as alternative available investments. Indications that this was not the case could be if the investment was not made on market terms or if it involved a type or degree of risk that was inappropriate for with-profits policyholders. In particular, there are doubts about the suitability of investments that give exposure to UK life insurance business risks, since policyholders are already exposed to such risks by virtue of their policy. An alternative implied term is that such investments should not be to the detriment of the with-profits fund (as distinct from policyholders) if the approach is that shareholders’ as well as policyholders’ interests in the fund should be taken into account.

5.2.4 Questions have been raised recently about ‘strategic investments’, in the equity or debt of a company with which the insurer's parent has a strategic connection or interest (Treasury Committee, 2008). The FSA (2007b) has confirmed that its rules permit ‘strategic investments’ to be held within a with-profits fund, provided this is fair to with-profits policyholders. This is consistent with the analysis above. FSA rule COBS20.2.32 concerning loans and guarantees, is also similar. The FSA (2010) is, however, alert to the issue that strategic investments, which may be illiquid, should not unfairly constrain payments to policyholders. More recently, it has proposed to amend its rules to require firms’ governing bodies to be satisfied, as far as they reasonably can be, that the purchase or retention of strategic investments is likely to have no adverse effect on the interests of with-profits policyholders (FSA, 2011b). The FSA's concerns appear to be consistent with the concerns expressed above.

5.2.5 The nature of strategic investments may be unclear. While the FSA (2003) decided to require insurers to disclose their policy on investments of this type, Sandler (Reference Sandler2002) wished to see more detail: there is a case for the regulator to require insurers to disclose the actual investments of this type, and their risks, in an annual report to policyholders. Such openness would be consistent with Lord Bingham's comments reported in appendix C3.

5.3 Management Service Companies

5.3.1 Several life insurers have entered into management services agreements (MSAs) with management services companies (MSCs) in their group, whereby the latter provide administration and sometimes acquisition and investment services for the life insurer, for a specified charge. These MSCs are usually 100% owned by the shareholders. The expenses used in calculating asset shares are then the charge paid by the insurer to the MSC; if the MSC incurs a lower (higher) expense than this, there is a profit (loss) to the MSC, i.e. to the shareholders.

5.3.2 Concerns have been expressed that MSCs may be a means of transferring profits from policyholders to shareholders (Brindley et al., Reference Brindley, Dumbreck, Thomson and Thompson1990, Reference Brindley, Nowell and Thomson1998) and may be inconsistent with TCF (Tuley, Reference Tuley2009).

5.3.3 Consider firstly an insurer that establishes an MSA. Regulators have been concerned that this could ‘milk’ the fund, taking profits from policyholders to shareholders: this was referred to in the Government Actuary's Department Insurance Supervision Manual, included in Parliamentary and Health Service Ombudsman (2008, part 4:674–676). That manual indicated that supervisors should look to ensure that the MSA charges were based on costs (although it does not explicitly rule out the MSC making a profit that would otherwise have gone to policyholders); and that there were no penalty clauses on termination of the agreement.

5.3.4 Establishing an MSA is a form of management action and the author would not expect it to be done to the detriment of customers, otherwise it would be unfair. He therefore deduces this is an implied term. To assess whether an MSA is in policyholders’ interests, it is necessary to consider the assets of the fund, after the costs it has incurred; and the risks it is exposed to.

5.3.5 Since the MSA bears the risk that expenses exceed the charges it makes, the fund is less exposed to risks. Everything else being equal, that is to policyholders’ benefit; and the benefit may be especially helpful in the case of large IT projects, for example. The reduction in risk can be particularly important for a fund that is weak or for a closed fund where declining business volumes may lead to an increase in unit costs.

5.3.6 Moving on to the assets of the fund, it no longer has to provide the resources, such as trained staff and systems, needed to supply the services now provided by the MSC. If the MSC pays the with-profits fund to acquire such resources, this enables the insurer to change operational resources into tangible assets, which it can invest in a suitable way. The assets of the fund would also be increased if the shareholders’ share of surplus is reduced to reflect the lower risks the fund is now exposed to.

5.3.7 A further reason for the assets to increase is if the charges made by the MSC are lower than the costs the fund was previously experiencing. This is because a 100% shareholder-owned MSC has clearer incentives to reduce expenses than a with-profits life insurer, where only (about) 10% of any expense saving accrues to the shareholders, the remainder of the gain benefiting policyholders’ payouts in a way that is not easily visible (Brindley et al., Reference Brindley, Nowell and Thomson1998; Sandler, Reference Sandler2002). An insurer in a group may also benefit from economies of scale that an MSA for the group may experience, leading to lower costs. Cost reductions can benefit shareholders and may also mean that the charges made to policyholders are lower than otherwise.

5.3.8 On the other hand, an MSC's charges may exceed the costs previously experienced by the fund as the MSC includes a profit loading, leading to the assets of the fund being lower than otherwise.

5.3.9 There are advantages and disadvantages of MSAs: in some circumstances they can be to policyholders’ benefit; in other cases they can be detrimental. Judgment is needed as to how valuable is the reduction in expense risk if this has to be balanced against a reduction in the assets available to the fund.

5.3.10 There are some reasons to think that policyholders would place only limited value on the reduction in expense risk. This is partly because the MSA will not remove all expense risks from policyholders, as the charges will generally depend on some cost or price index and will be re-set after a period. Further, policyholders are exposed to a variety of risks (e.g. market, mortality, expense) and from the perspective of integrated risk management it does not make sense for the policyholder to pay to mitigate each risk separately. Not only may some risks offset others, but the with-profits fund already provides some smoothing and guarantees (which helps explain why policyholders forgo 10% of distributed surplus). The gain to policyholders from additional security about expense risks may be small compared to other risks to their payouts, and it may not be worth paying the profit margin charged by the MSC, especially if that is benchmarked with margins quoted by an external service provider.

5.3.11 The FSA (2011b) proposes that an MSC should not charge in excess of its costs to a with-profits fund. Such charging is not necessarily detrimental to policyholders’ interests. For example, the MSC may have paid a sum to acquire the resources from the fund; the shareholders’ share of surplus may have been reduced below 10% to reflect the reduced risks to shareholders; there may have been efficiency savings arising from the new incentives; and the reduction in risk may be particularly valuable. However, this may well not be the case, and some would argue that the FSA's proposal is a reasonable response to the policyholder detriment it identified.

5.3.12 Once established, the operation of the MSA has to be not detrimental to policyholder interests to avoid being unfair, essentially another implied term. Ensuring that this is the case can be difficult. The FSA (2010) found no evidence that current charges were excessive or unfair across the sector, although some firms did not have clear benchmarks for costs or arrange timely reviews. The FSA report did not, however, indicate the basis for its conclusions: the author suggests that an indication that an MSC may be operated unfairly is if the MSC's prices increase more quickly than other prices. This is commonly the case, notwithstanding the expectation of productivity increases that should help offset salary increases. In closed funds, though, there may be diseconomies of small scale that justify higher price increases.

5.3.13 Further, where the MSC arranges for services to be provided from a firm external to the group, the author would not expect the MSC's charges to the insurer to exceed those charged by the external party. Unless some specific benefit from the MSC can be demonstrated, the insurer could contract direct with the external party.

5.4 Taxation

5.4.1 The taxation of life insurance companies is complex, and two main rates of tax are applied: the basic rate of income tax and the main corporation tax rate, known as the ‘policyholder rate’ and ‘shareholder rate’ respectively (HM Treasury & HMRC, 2010):

• – On pensions business:

  1. ○ investment income and capital gains are not subject to tax;

  2. ○ the part of the surplus attributable to shareholders is subject to tax at the shareholder rate of tax;*

• – On life insurance business:

  1. ○ the investment income and capital gains are subject to tax at the policyholder rate; and

  2. ○ the part of surplus attributable to shareholders is subject to further tax as the shareholder rate exceeds the policyholder rate of tax.*

* The asterisked parts of the tax liability have been referred to by some as “shareholders’ tax” (e.g. Treasury Committee, 2008), although this terminology should not necessarily be read as implying that the tax should be borne by shareholders.

5.4.2 It was the Finance Acts 1989 and 1990 that brought about changes so that shareholders’ tax became an issue:

• – Pensions business: previously, insurers typically computed their tax with provisions set in a way that meant firms paid little, if any, tax on their pension profits. From 1990, that was no longer possible, and transfers to shareholders on pension business generated a tax bill that had to be paid; and

• – Life insurance business: previously, there was only one tax rate, not two, applicable to life business; so that the further tax on shareholder transfers did not arise.

5.4.3 The question is who pays this shareholders’ tax? It could be paid from:

• – Shareholders (whose transfer of surplus is therefore reduced);

• – Asset shares (reducing bonuses and hence payouts to policyholders unless guaranteed benefits are greater; with a reduction in the transfer to shareholders of one-ninth of the reduction in the value of bonuses); or

• – The inherited estate (reducing its ability to support the fund and the potential distributions from it for policyholders and shareholders).

5.4.4 Certain features of the tax calculation can be highlighted:

• – The tax rate applied to shareholders’ tax is the ‘shareholders’ rate’, i.e. the normal rate of corporation tax payable by companies, not the rate applied to individual policyholders; and

• – The tax is payable as a direct result of and calculated on, the part of the surplus being transferred to shareholders (and is not payable by mutual life insurers).

5.4.5 The author suggests that the above features of the tax calculation mean it is reasonable to think that shareholders will pay tax at the normal corporation tax rate on their profits (the transfer of 10% of distributed surplus) and that shareholders’ tax will be paid by shareholders. This is therefore suggested as an implied term.

5.4.6 Further evidence in support of this is that, if the inherited estate is used to pay shareholders’ tax, then the inherited estate is eroded. This means reducing policyholders’ interests in future distributions or reattributions of the inherited estate, which is a concern of the FSA (see 4.2 above and FSA (2009)). If it is an implied term that shareholders’ tax should be paid by shareholders, a firm's decision to pay the tax from the inherited estate is to the detriment of policyholders and hence unfair.

5.4.7 There is also an argument that paying shareholders’ tax from the inherited estate is inconsistent with policyholders receiving not less than 90% of distributed surplus. This example is adapted from Which? (2008). Say £100 surplus is distributed, with the policyholders receiving £90 (in the form of bonuses), shareholders £10, and with shareholders’ tax of £1.80 (consistent with the rate indicated by Keefe, Bruyette & Woods (2008)). If that £1.80 is taken from the inherited estate, which is surplus earned in the past, then, in total, £101.80 surplus is being used to pay policyholders £90. This is only 88.4%. Which? obtained the opinion of counsel that it was difficult to see how this was consistent with the contractual obligations of 90:10 companies.

5.4.8 It might be argued that, in this example, the tax of £1.80 is a liability and therefore cannot be part of the surplus, which is not £101.80 but £100, of which policyholders receive £90 (90%) and shareholders £10. The logical consequence of this interpretation is that an increase in the shareholder rate of tax has no effect on what shareholders receive, which the author regards as counter-intuitive. The surplus was £100 before any decision was made as to whom it would be distributed. It is only because some is being distributed to shareholders that the tax liability is incurred and it is not clear why policyholders should be worse off as a result (i.e. receiving £90 out of £101.80 instead of out of £100). The author is therefore unconvinced by the argument which, in any event, does not invalidate the reasoning using the features of the tax calculation to derive the implied term that shareholders’ tax should be paid by shareholders.

5.4.9 Turning to who has paid the tax, Paul (Reference Paul1996) reported some limited research that concluded that it was more common for the tax to be paid by policyholders. A survey by Tillinghast Towers-Perrin (2001) found that shareholders often did not pay shareholders’ tax; in many cases it was paid from the inherited estate and, in some cases, from asset shares.

5.4.10 In 2004, the FSA introduced a rule (now rule COBS 20.2.20), that a firm may not attribute a shareholders’ tax liability to asset shares but may attribute it to the inherited estate if the firm can show that this is consistent with its established practice and it is explained in its PPFM.

5.4.11 The FSA rule is inconsistent with this paper's conclusion that paying shareholders’ tax from the inherited estate is inappropriate, and the author would argue that being established practice does not justify an inappropriate practice. There are also arguments that an explanation in the PPFM is irrelevant:

• – Many contracts began before PPFMs came into being in 2004;

• – PPFMs were drawn up by management without an opportunity for policyholders to comment;

• – most policyholders have not read the PPFM (they do not ordinarily receive a copy) and, if they have, may not understand the significance of the references to tax;

• – some policyholders will have read the customer-friendly version of the PPFM, but FSA rules do not require this version to explain that the inherited estate pays shareholders’ tax, if that is the case;

• – therefore, mentioning shareholders’ tax in the PPFM is not the same as disclosing this clearly to the policyholders concerned, which the FSA (2003), in its consultation, set as a requirement for the tax to be payable from the inherited estate.

5.4.12 It is useful to remind ourselves of the UTCCR: if the PPFM includes a term that the inherited estate will pay shareholders’ tax, that term was not individually negotiated, the firm has greater rights (to decide who pays tax) than the consumer, and there is consumer detriment. Mentioning the practice in the PPFM does not make it fair. It is also worth drawing attention again to Lord Bingham's comments about openness in appendix C3.

5.4.13 Indeed, given that the FSA believes it is unfair to pay shareholders’ tax from asset shares (which would reduce policyholders’ interests) it is not clear why the FSA believed it was fair to pay the tax from the inherited estate (which also depletes policyholders’ interests as they have an interest in the inherited estate).

5.4.14 In some cases, payment of shareholders’ tax by the inherited estate has been sanctioned by a scheme of financial management approved by the High Court (Office of Fair Trading, 2008). If the view is taken in future that paying shareholders’ tax from the inherited estate is inconsistent with the UTCCR then the directors of the life companies concerned may regard it as appropriate for the shareholders to pay shareholders’ tax.

5.4.15 The FSA's reasoning for its rule is not given in the consultation documents (FSA, 2003, 2004). It has referred to “preserving the shareholders’ position following a change in the tax law” (FSA, 2007b). It did not specify which change it is referring to, but the changes in the Finance Acts 1989 and 1990 were described by the government as “a revised approach to the allocation of investment income, capital gains and profits between shareholders and policy holders for various tax purposes” (Lilley, Reference Lilley1989). It was therefore a deliberate policy of government to bring about change. So it is unclear why the FSA should wish to preserve the previous tax position of shareholders.

5.4.16 The Treasury Committee (2008) concluded that the FSA's current rule “furthers shareholder interest to the detriment of policyholders” (page 4). This was consistent with the views, in their evidence in their report, of Clare Spottiswoode (the policyholder advocate in the Aviva reattribution) and Which? Insurers supported the FSA stance. Norwich Union's argument was that using the inherited estate may be the means to ensure that shareholders actually receive their 10% share of surplus. The author's argument above is that shareholders receive 10% gross but that using the inherited estate to pay shareholders’ tax so that the shareholders can receive 10% net is to customers’ detriment and unfair. The actuarial profession's (2008) evidence acknowledged the different views; a later paper by Tuley (Reference Tuley2009) indicated that using the inherited estate to pay shareholders’ tax may represent “sheltering shareholders from unwelcome tax changes”.

5.4.17 The FSA (2004) described its decision to allow shareholders’ tax to be paid from the inherited estate, in certain circumstances, as a ‘concession’ [to the industry]: this appears inconsistent with its responsibility to protect policyholders. More recently, the FSA (2011b) admitted that its rule has attracted considerable opposition, but pointed out that Solvency II will lead to tax changes (see HM Treasury & HMRC, 2010) although it is not yet clear what these will be. The FSA will return to this in a further consultation.

5.4.18 The FSA's (2011b) latest paper expresses concern about new business written in the expectation of a loss, which implies an erosion of the inherited estate and an adverse effect on with-profits policyholders’ interests as there is less money to distribute to them. The FSA therefore proposes to strengthen its rules and guidance in this area. The author suggests that this would rule out writing new business on the basis of shareholders’ tax being paid from the inherited estate. This would also have a benefit of ending the advantage available to with-profits insurers but not to others, which may distort competition in the market.

5.4.19 An additional point concerns pension policies where the insurer committed itself to providing a tax-free return, on the basis of marketing literature or other documents. That implies that shareholders’ tax will not be deducted from asset shares. Some firms did charge shareholders’ tax to pension policy asset shares before the FSA introduced new rules in 2004. The author considers that fairness (and meeting contractual obligations) requires such firms to declare bonuses such that its payouts to policyholders reflect a tax-free return including the period before 2004, and to have adequate provisions to achieve this.

5.5 The Impact of the Statutory Solvency Basis on Distributions

5.5.1 One further issue is discussed as it has been referred to in the actuarial literature for some time; its importance now is much reduced, though; and it may not apply at all following the implementation of Solvency II. This relates to shareholders receiving a share of distributed surplus that is one-ninth of the value of policyholders’ bonuses. Bonuses are a benefit payable in the future and a discounted value needs to be placed on the bonuses in order to calculate the transfer to shareholders.

5.5.2 One approach is to discount the policyholders’ bonuses at the gilt yield for the relevant duration of the policies concerned. This reflects the bonuses being guaranteed and is a fair valuation of the benefits.

5.5.3 The traditional approach of most insurers is to discount using the yield that they use in their statutory solvency valuation which, when using a net premium valuation, is lower than the yield on gilts as it has to be prudent. This leads to an increase in the amount the shareholders receive, compared to the position if the gilt yield was used.

5.5.4 This can be seen by considering an insurer that declares a 3% bonus on a guaranteed benefit of £1000. The cost of that bonus is 30/(1+i)ⁿ, where i is the rate of interest used, and n is the term from the valuation date to the maturity of the policy (say 10 years; and mortality is ignored). For example, in 1995, the 10-year gilt yield was 7.47%, and the cost of bonus on that basis was £30/(1.0747)¹⁰ = £14.60. However, the average rate of interest used in the statutory solvency valuation by four major proprietary with-profits life insurers was, for pensions business, 4.56%, giving a cost of bonus of £19.21. The surplus distributed to shareholders is one-ninth of the cost of bonus, i.e. £1.62 using the gilt yield, but £2.13 using the statutory solvency valuation rate of interest, i.e. 32% higher.

5.5.5 This ‘distortion’ (as it may be called) does not apply where no discounting is involved, namely most unitised with-profits business and for terminal bonuses on both traditional and unitised business.

5.5.6 If it is thought that the fair value approach is appropriate, then the distortion means that shareholders are, in ‘economic value’ terms, receiving more than 10% of surplus, which implies that the distortion is leading to customer detriment. Indeed, a number of authors have regarded typical practice as meaning that shareholders receive more than 10% of surplus as a result of valuing policyholders’ bonus at a low rate of interest: Bangert (Reference Bangert1973), Redington (Reference Redington1981), Brindley et al. (Reference Brindley, Cannon, Dumbreck, Thomson and Thompson1992) and Hare et al. (Reference Hare, Craske, Crispin, Desai, Dullaway, Earnshaw, Frankland, Jordan, Kerr, Manley, Mckenzie, Rae and Saker2004). Kennedy (Reference Kennedy1981:395) believed that shareholders were claiming an excessive level of surplus, saying, “Is it right, if the articles of association state that not more than 10% of the profits should go to shareholders, that we as actuaries should use an artificial method of calculation that gives them substantially more than 10%? Clearly that is not right”.

5.5.7 In some ways the traditional practice is understandable, as the surplus arises from having valued the liabilities in the statutory solvency valuation, where the regulations led insurers to use a discount rate lower than the yield on gilts. It is also a long-established and widespread practice, and can be regarded as a precedent for current practice. For some insurers it is referred to in the governing documents. When insurers make calculations for policyholders of projected payouts on policies, they take into account that the shareholders’ transfer is, in practice, worked out in this way. On the other hand, relying on past practice and documents of which the policyholder may be unaware is inconsistent with Lord Bingham's ‘openness’ (appendix C3).

5.5.8 The FSA (2003) proposed that proprietary with-profits life insurers should calculate the transfers to shareholders arising from reversionary bonus based on the long-term gilt yield (net of tax if appropriate). This would correct the distortion described above. The industry was however concerned that this may unduly reduce the value of shareholder interests (FSA, 2004). Therefore, the FSA made a concession that the transfer to shareholders could be calculated using the “low” rate in the statutory solvency valuation if the firm could show that this was consistent with its established practice and if it is explained in its PPFM. The FSA (2004) indicated that it continued to have reservations about the fairness of a method of calculating bonuses which can result in shareholders receiving more than the 10% of distributed surplus over time that they claim to receive.

5.5.9 The author's conclusion is that a policyholder or outside observer would not reasonably expect the shareholders’ share of surplus to be increased by the ‘distortion’ and that this represents customer detriment. A reasonable practice, to be consistent with dividing surplus 90:10 is, applying a fair value approach, to discount policyholders’ bonuses using a gilt yield (net of tax as appropriate). While the PPFM may indicate that the firm's practice is different, it is inappropriate to rely on this (as in section 5.4).

5.5.10 The distortion is, however, now much less than it was because:

• – reversionary bonus rates have reduced (in some cases to zero);

• – the average term to maturity of business in force has been reducing, with little new with-profits business being written: the discounting is therefore over a shorter period and the distortion correspondingly less;

• – the statutory solvency rules now permit a gross premium valuation to be used, in which the rate of interest is higher than in the net premium valuation method, leading to the distortion being reduced;

  • ○ with the ‘genuine’ shareholders’ share being 10–11% for conventional 10-year policies maturing now; and

• – most with-profits business now is unitised, where, in most cases, discounting does not arise and hence there is no distortion (FSA, 2004).

6.1 Introduction

6.1.1This section of the paper considers fairness when a proprietary with-profits life insurer carries out a reattribution of the inherited estate. Policyholders are given the option of accepting an offer of cash and/or increased benefits (policyholders’ incentive payment or ‘PIP’) in return for their giving up rights to future distributions from the inherited estate.

6.1.2One reattribution has been carried out under current FSA rules, involving the CGNU Life and CULAC funds of Aviva in 2009. A similar case in 2000, predating the rules, involved AXA. Prior to 2000, some insurers reached agreement with the regulator on how their inherited estate was to be divided between policyholders and shareholders (Smaller et al., Reference Smaller, Drury, George, O'Shea, Paul, Pountney, Rathbone, Simmons and Webb1996).

6.2 The Source of Inherited Estates and Possible Implications for Fairness

6.2.1One possible reason for the existence of inherited estates is that monies were injected by the shareholders at the outset (or later). Whether that is so is a matter for historical enquiry and the outcome in a particular case may be unclear. If it was the case, then (at least some of) the inherited estate does not represent profits as such: the fund could have been operated without (that part of) the inherited estate being treated as profits to be shared with policyholders, instead being regarded as ‘owned’ by the shareholders.

6.2.2On the other hand, capital injections may have been made without a clear basis that they represented shareholders’ assets, with policyholders believing that any such monies formed part of the fund, contributing to surplus that is shared in the usual way between policyholders and shareholders. The following discussion is based on this assumption, or that there is no demonstrable injection of capital by shareholders.

6.2.3There are reasons for thinking that an inherited estate may well be, largely or wholly, a result of past with-profits policyholders receiving less than asset shares. For example, prudence may have led to payouts being smoothed below asset shares when the latter were thought to be inflated as a result of strong investment returns on equities over a long period. Insurers may have also wished to err on the side of under-paying if their systems for asset shares or bonus rates were inadequate (Smaller et al., Reference Smaller, Drury, George, O'Shea, Paul, Pountney, Rathbone, Simmons and Webb1996).

6.2.4Some may argue that fairness requires policyholders in a reattribution to receive a PIP that reflects a 90% share of the inherited estate because that reflects what policyholders have contributed. This paper instead analyses the position from what are the terms, including implied terms, of the with-profits contracts.

6.3 Valuation of the Inherited Estate for a Reattribution

6.3.1A reattribution is essentially a transaction whereby an inherited estate in which policyholders and shareholders have rights becomes owned by the shareholders. On the face of it, the market-consistent value of the inherited estate is needed as the basis for the transaction. A good starting point is the insurer's published realistic balance sheet in the FSA Returns. However, it is implicit that the calculation of the inherited estate for use in a reattribution has to reflect the characteristics of that transaction in order to reflect the value of the long-term business fund assets after the insurer has settled its liabilities. The author suggests the following adjustments to the value in the realistic balance sheet:

1. 1) Any prudence in the realistic balance sheet calculations should be eliminated;

2. 2) Staff scheme pension costs should be valued realistically;

3. 3) Management actions should be included as will apply post-reattribution. For example, reducing the equity backing ratio can lead to a reduction in liabilities as guarantees are less likely to bite, and taking such actions may be more likely given the changed incentives post-reattribution. An insurer may later wish to change its management actions from those specified, for example because of low solvency brought on by adverse investment conditions. This is understandable. However, the term implied from section 5.1 would mean that this could not be done in a way that advantages shareholders over policyholders;

4. 4) Liabilities should be valued on a “settlement” basis, i.e. what the insurer expects to pay in settling its liabilities. If, for example, the insurer's expenses are expected to reduce, that should be reflected in the calculation;

5. 5) Goodwill (and any other inadmissible assets) should be included as an asset. Tuley (Reference Tuley2009) mentioned that a longstanding fund, with administration systems, marketing connections and brand value can have a goodwill value;

6. 6) The liabilities should be increased by the value to policyholders of the security provided against risks not included in the realistic value of liabilities. Three points are mentioned:

• – The put option prices used to value guarantees, etc may not fully reflect the value being provided; for example, the prices relate to an index of assets, whereas the actual assets will differ;

• – There may be some expenses that would have been met from the inherited estate in the past, e.g. the cost of administrative errors; if they arise in the future and are to be paid from the (then shareholders’) inherited estate, that is a benefit of value to policyholders; and

• – Insurance and operational risks mean that mortality/morbidity and expense (and tax) experience may be worse than assumed, increasing the value of guarantees. Placing a valuation on this benefit is difficult. However, it needs to reflect not only the cost to insurers of supplying the security but also the value placed on it by policyholders (a balancing of supply and demand factors).

6.3.2Note that an insurer's liabilities are commitments related to past and not to future events. This means that the inherited estate should not be subject to a deduction for future development costs or for any expected new business subsidies. This differs from FSA's past practice in allowing a deduction from the inherited estate for non-market risks not charged for (LECG, 2009c). However, it is right to take into account that the goodwill asset will be affected by future development costs and new business pricing.

6.4 Accuracy of the Inherited Estate Calculation

6.4.1Policyholders cannot be expected to give up future rights to distributions if the size of the inherited estate is unclear. However, the author suggests that it is far from easy to calculate the inherited estate as adjusted above. Indeed, considering just the inherited estate as reported in the realistic balance sheet, there are questions about the robustness of the data, the modelling and also the valuation of options and guarantees (O'Brien, Reference O'Brien2009b). Clearly, it is important that the calculation of the inherited estate is fit for purpose.

6.5 Interests of Policyholders and Shareholders in the Inherited Estate

6.5.1 The question of what rights policyholders have in the inherited estate has been the subject of heated discussion in the actuarial profession. The Smaller et al. (Reference Smaller, Drury, George, O'Shea, Paul, Pountney, Rathbone, Simmons and Webb1996) working party reported the views of two groups.

6.5.2 Some members of the working party felt that the ownership of the inherited estate should be in the same proportion as for distributed surplus unless there were any special circumstances. The following points were felt to be relevant:

• – If payouts were less than asset shares in the past, it would be wrong for the shareholders to gain from having a higher entitlement in the inherited estate compared to distributions of surplus.

• – If earlier generations of policyholders had made contributions to the inherited estate that were now regarded as excessive, it was difficult to see any a priori reason for the excess then being distributed in a different proportion from current profits; and

• – Whilst distributing some of the inherited estate to current policyholders means such policyholders gain, this should be seen in the context of it not being unusual for insurers to smooth returns by transferring gains and losses between generations.

6.5.3 Others felt that it was unsound to rely on the arbitrary 90:10 ratio to distribute the inherited estate without proper regard to the origin and growth of the inherited estate. In particular, they commented:

• – Equity was not served by transferring undistributed profits from past terminated policies to those now in force and others yet to be written;

• – There was often no identity of interest between the policies that generated the inherited estate and those now in force;

• – Transfers of profit between generations should not be made, except to achieve an acceptable degree of smoothing; and

• – With-profits policyholders, as a class, cannot have a right to the inherited estate, because there is no requirement to distribute it.

6.5.4 It is possible to consider interests in the inherited estate if a reattribution does not take place, by considering the circumstances in which the inherited estate would be distributed. These are if:

• – the fund closes to new business;

• – the fund has excess capital, which is distributed; and

• – the fund is wound up.

6.5.5 There may be some part of the inherited estate that is not expected to be distributed: the author refers to this as the ‘residual’.

6.5.6 If the fund closes to new business, the firm has to prepare a run off-plan, and FSA rule COBS20.2.17 means that, in a 90:10 fund, policyholders will receive not less than 90% of the inherited estate following closure.

6.5.7 Whether a fund closes depends on management policy. This is more likely than otherwise as a result of the reduced demand for with-profits products, and also as many funds have weakened. Some funds have written new business at a loss to the inherited estate, which may be a precursor to closure even if the fund is strong. The author's view is that each fund must have a non-zero probability of closing at some date in the future.

6.5.8 FSA rule COBS20.2.21 requires with-profits insurers to consider, annually, whether they have an excess surplus, which may lead to a distribution of some or all of that excess. The FSA (2007b, 2008a) has indicated that where a firm has excess surplus it would ordinarily require a distribution, where 90% is allocated to policyholders; more recently, it has proposed that reattribution will not be permitted for excess surplus (FSA, 2011b). Even where a fund has no excess surplus, it has the potential for excess surplus to arise at some future point, as a result of the experience of the fund, with a distribution following.

6.5.9 This demonstrates that policyholders have a contingent interest in the inherited estate, depending on future distributions which can arise upon closure or there being excess surplus. The FSA (2008a, 2009, 2011b) confirmed that policyholders have a contingent claim on the assets of the inherited estate.

6.5.10 An insurer may write new business where new with-profits policyholders do not have rights to future distributions from the inherited estate. If so, the author suggests that it is an implied term that the inherited estate should begin to be distributed 90:10 to current with-profits policyholders and shareholders. Otherwise, some inherited estate assets would remain after the last policyholder with an interest in the inherited estate had left, leaving the whole of the residual inherited estate to the shareholders. This is inconsistent with distributing on a 90:10 basis.

6.6 Interests of Policyholders and Shareholders in a Reattribution

6.6.1 Policyholder and shareholder interests are considered by first assessing what would be the flows of future surplus distribution in future scenarios, including the fund being closed or having excess surplus. If the probability of each scenario is estimated, it is possible to calculate the expected flows of surplus to the constituent parties, namely:

• – current with-profits policyholders;

• – future with-profits policyholders; and

• – shareholders.

6.6.2 This exercise is complex. For example, it involves probabilities of the fund closing in each future year, which depends on future management decisions. Management may say they have no intention to close the fund, but of course a (different) future management may act differently.

6.6.3 Financial modelling can be used to assess the likelihood of excess surplus. There are again uncertainties, both as regards how satisfactory the model is, and the dependence on management actions, such as the amount of new business written, the firm's investment strategy and its decisions on when surplus is excessive.

6.6.4 For distributions arising following closure or excess surplus, the share of those who are policyholders at the time is 90%, in accordance with FSA rules.

6.6.5 There may be a ‘residual’ part of the inherited estate, which is not expected to be distributed as a consequence of closure or excess surplus. Consider this in the context of the three basic with-profits principles. The residual is part of the surplus of the fund; although earned previously, this is not a barrier to policyholders gaining from it as with-profits business involves inter-generational transfers. And as surplus, including inter-generational transfers, is divided 90:10, the author concludes that it is implied that policyholder and shareholder interests in the residual are also 90:10.

6.6.6 The author therefore concludes that policyholders have an interest in 90% of the distributions from the inherited estate and a 90% interest in the residual.

6.7 The Value Expected to be Distributed to Future Policyholders and the Residual

6.7.1 Say there is an inherited estate of 100. Assume it is expected to be distributed: 27 to current with-profits policyholders, 36 to future with-profits policyholders, 7 to shareholders and 30 is the residual. Say the insurer offers a PIP of 35. Current policyholders may be inclined to accept it as they are only expecting 27. That gives the shareholders 65. In other words, shareholders can gain the share expected to be distributed to future policyholders, who have no choice in a reattribution, and they can also claim the residual.

6.7.2 Alternatively, revert to the three basic with-profits principles. Surplus is shared 90:10. Given that there are inter-generational transfers of surplus, surplus may be retained for the future, and 90:10 still applies. The author therefore derives a fourth principle, the ‘distribution-independence principle’: namely that the proportionate interests of shareholders and policyholders are unaffected by whether surplus is actually distributed in any year. This is as in the Ministerial statement of 1995 (see section 6.11): if, in any year, surplus is not distributed, but is retained in the fund, the policyholders, as a class, retain their interest in that element of the fund. Since, it is assumed, the inherited estate is built up from past undistributed surpluses, the interest of policyholders in the inherited estate is 90%.

6.7.3 This fourth principle underlies the FSA rule that with-profits policyholders are entitled to 90% of the inherited estate if a fund closes. Accepting the principle means that the with-profits policyholders at the time a fund closes or is reattributed, gain from past surpluses in a way they were not expecting. But, in the author's view, this is not inappropriate; indeed, it is necessary to preserve the 90:10 division of profits. The position is also similar to a demutualization, where those who are policyholders at that point in time share the inherited estate.

6.7.4 Given that a reattribution buys out policyholders’ interests, if the price paid (the PIP) is to reflect those interests, it will reflect the value of 90% of expected future distributions from the inherited estate and 90% of the residual. The author's view is that, if a PIP did not reflect those interests, it would be detrimental to customers’ interests, i.e. unfair.

6.8 Imbalances of Rights and Obligations

6.8.1 A reattribution is effectively an option to the insurer to offer policyholders the option to accept a PIP instead of future distributions from the inherited estate. However, the way a reattribution is carried out involves an imbalance in the rights and obligations of the parties:

• – The firm can choose the timing of the offer to suit itself;

• – The firm has more information about its current finances and potential future actions than policyholders;

• – Most policyholders have limited time, information and understanding to consider the PIP offer, whereas the firm can ensure it has the necessary resources;

• – The firm is the body that makes decisions, and can make assumptions about its plans for new business and for investment, bonuses and other matters, with the ability to change its approach to new business and so on after the reattribution;

• – Policyholders’ information about the appropriateness of the PIP is limited as only one firm (the insurer) is offering to buy out interests in the inherited estate; policyholders would gain if it were feasible for other firms to buy out their interests, creating a competitive bidding process; and

• – The firm has a long-term perspective, where all future cashflows are relevant to its calculation, whereas a policyholder may only consider cashflows over the lifetime of his or her policy.

6.8.2 These imbalances may lead to the firm offering a lower PIP than section 6.7 suggested was implied to avoid unfairness. This possibility may have other consequences since shareholders receive 10% of profits in ‘ordinary’ distributions of surplus, if the fund is closed and if the fund has excess surplus. If shareholders could receive more than 10% in a reattribution, insurers have an incentive to reduce ‘ordinary’ distributions of surplus (Sandler, Reference Sandler2002); to reattribute before closure; and to increase the amount of surplus that they regard as not excessive. These points are considered further below.

6.8.3 First, an insurer could reduce ‘ordinary’ distributions, paying policyholders a lower proportion of asset share than otherwise and/or imposing high charges for guarantees, etc. This was a concern of Clay et al. (Reference Clay, Frankland, Horn, Hylands, Johnson, Kerry, Lister and Loseby2001): they referred to the criticism made that policyholders may be receiving less than ‘fair share’ so that insurers may build up their inherited estate in order to benefit their shareholders. However, FSA rules on payouts and charges have been tightened since Clay et al.'s paper, so the significance of the issue has reduced.

6.8.4 Secondly, an insurer could claim that it is unlikely to close and could take a number of actions to avoid closure, enabling a reattribution to take place, the policyholders then receiving less than 90% of the inherited estate. The firm could:

• – deny that closure was a possibility;

• – write small amounts of new business even if at high cost;

• – use the inherited estate to subsidise new business (the three funds with a reattribution project in 2007 all wrote new business at a loss to the inherited estate in 2006 and 2007, according to data in FSA Returns); however, in the future, insurers will be constrained by the FSA's (2008a) statement about not allowing unreasonable subsidisation, and further if the FSA's (2011b) proposals to tighten its rules in this area are implemented; or

• – merge it with an open fund when it would not otherwise have done so.

6.8.5 Thirdly, an insurer could claim that it needed a large amount of capital, thereby reducing the probability of having excess surplus, by:

• – increasing its new business forecasts to increase the capital it requires;

• – subsidising new business, if permitted by the regulator, which would also add to the capital it claims to require;

• – being prepared to take more risks, for example by adopting a more risky investment strategy; and claiming it needs large amounts of capital to offset those risks; and

• – carrying out its calculations on a more prudent basis than otherwise.

6.8.6 If the insurer increases its new business forecasts, this not only reduces the probability of excess surplus; it also increases the share of the inherited estate expected to be payable to future policyholders. Under current FSA rules, the insurer can gain from then offering a lower PIP. Management can subsequently influence the actual new business which, if it is less than forecast, means excess surplus arises more quickly, to shareholders’ benefit after a reattribution. It is not surprising that, in the AXA and Aviva reattributions, there were disagreements between the firm and others about the new business forecasts: see Jenkins (Reference Jenkins2000), Aviva UK Life (2009d) and LECG (2009b).

6.8.7 The role of the policyholder advocate addresses the imbalance between policyholders and the firm to some extent. However, the policyholder advocate in the Aviva case, Clare Spottiswoode, said that intergenerational transfers skewed bargaining power in favour of the insurer as voting policyholders would consider their expectation of receiving a distribution from the inherited estate whereas the firm would look further ahead. While she could look to influence the PIP offer, her negotiating power was weak (Spottiswoode, Reference Spottiswoode2009a).

6.8.8 In conclusion, policyholders are in a weak position in a reattribution and firms can increase shareholder value at the expense of policyholders. To permit a firm to offer a PIP that does not reflect 90% of the expected distributions of surplus in the inherited estate and 90% of the ‘residual’ is unfair in the author's view as this takes advantage of the imbalance to give policyholders less than was derived as is implied in section 6.7.

6.8.9 It is worth adding a reference to the CPUTR, where Regulation 3(3)(b) defines a practice as unfair if it “materially distorts or is likely to materially distort the economic behaviour of the typical consumer with regard to the product”. The author suggests that a reattribution may be unfair in that context as it gives an incentive to policyholders to accept a proposal that is arguably inconsistent with the contract.

6.8.10 Therefore, while it is true that a policyholder has the choice of not accepting a reattribution offer, the author would expect the regulator to intervene if the offer was less than was derived as is implied in section 6.7.

6.9 Liquidity and Certainty Benefits

6.9.1 Which? (2008) argued that shareholders should not be able to obtain more than 10% of the inherited estate. However, the author does not conclude that a PIP, to be fair, must equal 90% of the inherited estate. Some policyholders may feel that they are gaining by receiving a specific PIP instead of uncertain future distributions, especially if the PIP is in cash form. There is value in that liquidity and certainty: for example, if the policyholder has a more liquid asset, he or she may be able to save money as a result of borrowing less. Indeed, given that the insurer is the party that manages the inherited estate, there can be merit in the uncertainty and illiquidity being borne by the insurer rather than policyholders. Aviva UK Life (2009c) referred to the ‘value unlocked by the reattribution’.

6.9.2 An illustration of the liquidity benefit is that the insurer would incur costs in raising external cash to fund the PIP. In the Aviva reattribution, LECG's (2009a) calculation allowed for transaction costs that might be incurred by shareholders, if funding for the PIP was raised externally, of 3.5% of the aggregate PIP.

6.9.3 Policyholders may also prefer a certain payment rather than a right to uncertain future cashflows. Consider the fair (market) value of this benefit of certainty, perhaps by envisaging a policyholder's interests in the inherited estate as saleable if there were a competitive environment. As the inherited estate is very largely tangible assets that have a market value, the author suggests the value of the certainty would be relatively low.

6.9.4 Therefore, if the aggregate PIP is to reflect 90% of the distributions from the inherited estate and 90% of the residual, the PIP can be less than 90% of the inherited estate in order to reflect the fair value of the liquidity and certainty benefits that policyholders gain. In practice, there is not a competitive environment to value these benefits. However, the author suggests that the deduction from 90% to reflect these benefits, on a fair value basis, would be quite modest.

6.9.5 The author's view is that an offer of a lesser PIP is taking advantage of the weakness of policyholders and is unfair.

6.10 Frictional Costs

6.10.1 In practice, there are significant frictional costs: expenses and tax. The expenses of a reattribution exercise can be expected to be substantial. In the Aviva reattribution, LECG (2009a) used £147 m as ‘allowable costs’ in the calculations; the full costs were higher than this.

6.10.2 There is also a major tax disadvantage: the insurer does not gain tax relief on the cash PIP it pays to policyholders, but it is taxed on the investment return on the inherited estate reattributed to shareholders and on distributions therefrom (Aviva UK Life, 2009b). If the PIP was in non-cash form, tax relief can be claimed, but such an offer does not provide a liquidity benefit to policyholders and is likely to be of limited appeal. The insurer also needs to consider tax on the investment earnings of the inherited estate assets post-reattribution; depending on the circumstances, this may create an advantage or disadvantage.

6.10.3 The author's view is that these frictional costs mean it is highly unlikely that a reattribution can be carried out in a way that gives shareholders and policyholders a satisfactory deal and in accordance with what the author believes are the implied terms of with-profits contracts.

6.11 Regulatory Action

6.11.1 It is natural for regulators to be closely involved in this subject. The Ministerial statement issued by the Department of Trade and Industry in 1995 is (see Smaller et al., Reference Smaller, Drury, George, O'Shea, Paul, Pountney, Rathbone, Simmons and Webb1996:1282–3):

6.11.2 The Ministerial statement is consistent with the distribution-independence principle. It preserves the rights of policyholders as a class. The FSA (2008a) confirmed that it abided by the ministerial statement fully.

6.11.3 Notwithstanding this, FSA is prepared to consider a reattribution where the shareholders would receive more than 10% of the inherited estate. The FSA (2008b: Ev. 78) stated that it decides if the proposal is fair “compared with policyholders awaiting a potential future 90:10 distribution”, noting that the inherited estate may not become available for distribution. The author's concern is that this does not satisfactorily address equity between shareholders and policyholders, as if policyholders do not receive a share of the inherited estate because there is no distribution then, in accordance with the 90:10 principle, neither should shareholders.

6.11.4 The FSA does have a number of other tests including, in particular, the rate of return that the reattribution provides for shareholders. Two particular points are significant. First, the FSA calculate the rate of return taking into account the firm's new business plans. However, Spottiswoode (Reference Spottiswoode2009b) indicated that the firm should take into account, when deciding on the terms of new business, the effect on shareholder value of a deferral of distributions from the inherited estate; the author therefore suggests the rate of return should be assessed ignoring the effect of new business. The second point is more basic. Given the high frictional costs of a reattribution, the author would not expect a reattribution to give good value for policyholders and shareholders. If the rate of return were satisfactory for shareholders, this means it is unreasonably low for policyholders.

6.11.5 The AXA and Aviva reattributions were approved through a court process, though views differ on whether the judgments were satisfactory. Mr Justice Norris indicated that he judged the Aviva case on the law as it stood rather than the law that Ms Spottiswoode, the policyholder advocate, would prefer it to be ([2009] EWHC 2521 (Ch.)). This paper instead judges the fairness of a reattribution without taking FSA rules as given.

6.11.6 The author's view, following the analysis in section 6, is that the FSA should not regard a reattribution as fair unless:

1. (a) the calculation of the inherited estate for use in a reattribution reflects the characteristics of that transaction; and

2. (b) the aggregate PIP reflects the value of 90% of expected future distributions from the inherited estate and 90% of the residual, and can reflect the liquidity and certainty benefits to policyholders.

6.12 Conclusion

6.12.1 A firm proposing a reattribution may regard it as fair for a policyholder to choose the PIP if it exceeds the present value of expected distributions, to that policyholder, from the inherited estate. That is an understandable position: a policyholder may regard it as unlikely that he or she will receive much from potential distributions and it can be welcome to receive an (unexpected) PIP. In the AXA and Aviva reattributions, the vast majority of policyholders chose to receive the PIP. However, the author believes that such a position is inappropriate as it does not address the arguments made above, in the context of the rights of with-profits policyholders.

6.12.2 A reattribution could be judged more widely, such as whether the aggregate PIP is more than some specified proportion of the inherited estate or whether the aggregate PIP includes not only the expected value of distributions but also some specified proportion of the residual not expected to be distributed. Ad hoc rules on these lines can be designed and they may be helpful in evaluating the PIP. However, the author suggests that they do not resolve meeting the rights of policyholders, as considered above.

6.12.3 This section has considered a reattribution in general terms. There are other important issues, in particular the allocation of the aggregate PIP between individual policyholders, and ensuring that policyholders who do not elect to take the PIP are treated fairly. Judging the fairness of a reattribution must take into account the actual circumstances of the fund. The reattribution scheme may also have additional provisions. For example, it may provide for some adjustment if the fund were closed to new business in a specified period after the reattribution, or there may be constraints on the ability of shareholders to access the inherited estate after the reattribution. However, the author's view is that such provisions are highly unlikely to lead to an outcome that he regards as fair in accordance with the implied terms of with-profits contracts.